Until 1989, the American Association for Applied Linguistics (AAAL) could have been viewed as an interest group of the Linguistics Society of America (LSA); AAAL met in two designated meeting rooms as a subsection of the LSA conference. In 1991, I was asked to organize the first independent meeting of AAAL in New York City, with the help of Nathalie Bailey as local chair. With the planning committee, we made several important changes that would take effect in the first independent AAAL conference: we would have many more sessions than could be accommodated in two rooms, we would invite colloquia as well as individual papers, and we would go out of our way to recruit international presenters, even though our name was the american Association for Applied Linguistics. I remember that first independent AAAL conference as a resounding success, with 400 enthusiastic participants.

We are presently participating in the twenty-second independent AAAL conference. Since I have been given a podium, I am going to take the opportunity to revisit a fundamental question that we dodged in our planning meetings for the 1991 conference, namely: How will AAAL define itself? What is applied linguistics? We dodged this question back then; we refused to define ourselves. Rather, we chose what I remember as Bernard Spolsky's solution; he said simply something like: ‘applied linguistics is what we do’. We would define applied linguistics as what we did, rather than what we said. Now, 22 years later, what exactly is that? I want to frame this personally, so: what is it that I do, what have I done, as an applied linguist?

An outsider might think linguists create theory, and applied linguists apply that theory to practice. But as we know, it is not that simple. Like Spolsky, I will not address this question head on, at least at first. Rather, I would like to begin with an analogy to a related field: the field of medicine.

It is obvious that there is variety in what different kinds of doctors do. My stereotype is that doctors work in clinics and hospitals and university medical schools; when I think of medical researchers in particular, I imagine them in white coats, surrounded by state-of-the-art computers, instruments, and medical labs, and supported by sophisticated staff. I imagine them in an antiseptic, modern, and protected world where the reward systems are set up to encourage medical specialization, diagnostic and research skills that can result in publication, and brilliance in the construction, evaluation, and defense of their positions in the research literature. One interesting characteristic of medical research is that it is not just large-scale experimental studies that get published; unusual case studies of individuals are also highly valued. Finally, doctors seem to operate as the centers of their professional worlds: patients and research participants come to them, rather than vice versa.

A very interesting, different, kind of medical practitioner is the médecin sans frontières (MSF); one English translation I have seen is ‘doctors without borders’. I think it is the name I like: it appeals to me, the idea of professionals without borders – assuming that's even realistically possible. I have never actually met an MSF, of course, and what I know is based on some news stories and what I see on their website. The first noticeable thing about the website is what they say the MSFs do. According to the mission statement, MSFs ‘deliver emergency aid to people affected by armed conflict, epidemics, healthcare exclusion and natural or man-made disasters’ (Médecins sans Frontières 2012). So these doctors without borders do not stay put in professional territories. It is not that they have no borders; they cross them: national borders and social boundaries. They transport themselves out of their clinics, hospitals, and medical schools to deliver care to people living in remote parts of the world who apparently would otherwise have no medical care at all. They leave a protected and perhaps antiseptic world of clinics and research labs and deliberately go into places that are not antiseptic – and may be fraught with danger. They deliver their skills to people who are having problems their skills can remediate. Does the fact that they are border-crossers make them perfect? I am sure it doesn't. Is it possible they may bring a colonizing mentality when they practice in the developing world? Possibly. But I give them credit for taking risks to try to apply their skills in unexpected settings, and across borders that might be barriers: national, cultural, economic, racial, linguistic, religious, educational. As they do that, they put themselves in places where they have tremendous opportunities to receive as well as give, learn as well as teach, be healed as well as heal.

It appears the reward systems of these different types of medical professionals are very different. Of course researchers and academicians in this, as in every profession, are rewarded for specializing. But the doctors without borders must cultivate breadth of medical knowledge in order to cope with the unpredictability of disease in unfamiliar local cultural contexts. Where medical researchers and academicians are rewarded by a kind of general fame and status – the professional recognition that comes from the publication of research papers – doctors without borders have far more local rewards, such as the improved health of an individual or the saving of a life. Where the academic world is adversarial and aligned to one or another theoretical position, MSFs are supposed to be guided by principles of neutrality and impartiality so as to improve the quality of local human life. What is interesting is that this impartiality explicitly aligns itself with advocacy. As the MSF website further states: ‘The principles of impartiality and neutrality are not synonymous with silence. When MSF witnesses extreme acts of violence against individuals or groups, the organisation may speak out publicly’.

So: two kinds of doctor. Is one kind better than the other? No. It is clear not only that both kinds of medical specialists are necessary, but that in some way they need each other. It is essential for both kinds of medical practitioner to exist and to have a good two-way flow of information between them. Theory and medical knowledge shouldn't just flow from the high-tech Western medical school in one direction, out to the developing world. Knowledge and data should also flow from the developing world back in the opposite direction: as when, for example, a new virus emerges in a local patient, or when an MSF learns from local practitioners about a folk medicine or local procedure that has obvious healing powers.

So what does this brief exploration of the MSFs do for us? Every analogy breaks down if you push it too far, of course. But can we see enough resemblance to identify similar ‘types’ in applied linguistics? I ask myself, who do I, an applied linguist for more than 30 years, most resemble: the academic researcher or the doctor without borders? First, I'd like to define the idea of ‘border’ or ‘boundary’ for us culturally, in terms of community and community norms. Borders can be literal, geographical boundaries between urban sub-groups, or between continents. Communities agree on language, so boundaries between communities usually demarcate linguistic differences. Boundaries can be political, as between Republicans and Democrats; disciplinary, as between psychologists and anthropologists; religious or economic or gender-based, and so on. Within an academic community's borders, members of that community view themselves as sharing academic goals, assumptions, practices, genres, terminology, rhetorical style – in short, sharing a culture. When a member begins to live or work in a group with different assumptions, practices, beliefs, or cultures, that member has crossed a border. No human being who lives and works in a community actually exists without borders; what we can do is cross borders, move into and out of different communities. So the question I am asking is, how much do we ‘do applied linguistics’ within an applied linguistics community that shares goals, norms, and assumptions, and how much does our doing of applied linguistics lead us in some way into other communities or subcultural sets of norms and assumptions where we can both apply our knowledge and learn something new?

Having framed my question in this way, I can now begin to list names of applied linguists who I see as crossing different kinds of borders in their work, and one name leads to another: Dennis Preston, who gathers data in places like truck stops and interviews ‘folk’ (meaning ‘not academics’) to model their mental maps showing where they think Americans ‘talk funny’; Dick Tucker, who with Wally Lambert brought a revolutionary new approach – language immersion education – to the world's attention; Jane Zuengler, Courtney Cazden, Patsy Duff, and others who enter classrooms with the respect of anthropologists entering villages in unknown cultures. Tere Pica, who asked unexpected questions about learning in classrooms, and mentored generations of graduate students to be border-crossers. Dick Allwright, who crossed unspoken boundaries in classrooms when he developed Exploratory Practice to put power into the hands of teachers in development; this challenged the norms of language teacher education by showing teachers and students how to carry out their own research to improve their own lives in the classroom. Bonny Norton's highly creative research, which challenges dominant theories by asserting the central impact of learners’ identity and imagined community on their investment in language use and language learning. Richard Young, whose study of discursive practice and interactive competence (including nonverbal communication) crosses multiple disciplinary boundaries, languages and cultures in a single stroke. And finally, my own M.A. ESL students, whose unexpected questions and curiosities lead them to unexpected discoveries about the way language works in the world.

Once you get started you can see that our field contains many examples and role models who are crossers of borders. They begin with an applied linguistics community's set of analytical tools, norms, and assumptions, and then move into other communities and spaces, or borrow from other disciplines, or create other teacher–student role relationships. Some may publish and take on an academic/researcher role, but this may not be essential for all of them, or as important to them as being able to go into local settings to use their skills to solve puzzles or problems. And in doing this, they learn themselves, learn facts, processes, new viewpoints, and explanations for the way humans use and process multiple languages that would have been utterly inaccessible to them if they'd stayed ‘inside the borders’ of normative academia. And their learning has benefitted the entire field by showing us new possibilities for theory and research.

I cannot voice narratives for the experiences of these other colleagues, but I can voice my own. What I have come to see is that I have been, and am, to at least some extent, doing applied linguistics both inside and outside the norms of my academic community, and have always felt tension in that duality. I have also come to realize how relatively easy it is for me to be rewarded in the academic researcher role – as a person who operates within the norms, or the expectations, of the discipline, publishing papers and books about theory and research – and how difficult it is to be valued and rewarded as one who crosses those borders, boundaries of academic units, discipline, culture, language, economic privilege, rank, or race, to apply knowledge about language to improve, at a local level, the lives of specific people on the other side, and to learn from them in the process.

The area of applied linguistics I know best is second language acquisition (SLA) research and second language teacher education. It seems clear at a conference like this how valued and rewarded SLA researchers and theorists are. Presentations usually orient to one or another theory, and present research data; and at the publishers’ exhibits, the SLA books most on display focus on theory, research studies, and research syntheses; presentations and publishers represent a world of abstract ideas where we align ourselves, often in adversarial ways, with one or another SLA theory. We can be very adversarial in promoting, comparing, supporting, or disproving a variety of universalistic principles, theories, and claims about the way humans become multilingual. I have my share of papers and books represented in this world of adversarial ideas. But I am not comfortable leaving it at that.

I think the field of SLA research is endangered when we over-value these norms of academic theory, research, and publication at the expense of different kinds of ‘border-crossing’ activities. For example, dominant norms lead us to gather SLA data from a surprisingly small, and undoubtedly unrepresentative, subset of the world-wide population of language learners – we focus on those in K-12 or post-secondary classrooms. As Bigelow & Tarone (Reference Bigelow and Tarone2004) pointed out eight years ago, SLA theories claiming to model the universal human capacity for L2 learning are based almost exclusively on data that are not representatively gathered. SLA researchers gather that data from a small subset of humans, the highly educated second language learners in K-16 educational settings in the developed world. Until recently there was no SLA publication of oral learner language data elicited from low-literate or non-literate adult L2 learners. And to this day, we have almost no data on SLA from multilingual learners in the developing world, from continents like Africa. I recently had the chance to read an eye-opening anonymous manuscript written by a scholar in Africa, who raised serious questions about whether the assumptions of SLA academic endeavor apply at all in that continent (Anonymous 2011). Specifically: What is ‘second language acquisition’ when a learner has varying levels of mastery of six or seven different languages, has completely lost mastery of her first language, and has literacy in none? Nor is Africa the only large area of the world where it's highly likely that there are learners that SLA theories do not explain or predict. In 1970 Jane Hill claimed there was work in anthropology describing adult learners in a remote part of Brazil who were successfully acquiring second languages after the critical period; to this day we've had no border-crossing SLA researcher actually cross that geographical and social border to gather data among the Vaupes people to test this claim. In medicine, there are some folks who leave academia in the developed world and put themselves in locations where they might encounter new kinds of data, and set up mechanisms for two-way exchange of such information – in SLA research, we need more such folks.

Nor are the borders that need crossing in the field of SLA just geographic or related to poverty or educational inequity. Some of the most important borders I struggle with are role-related: for example, one such border relates to the power structure in my own graduate classrooms. Why is it that when I am teaching teachers or graduate students in introductory survey courses about SLA research, I lecture so much? In the past, I have used a predominantly one-way transmission model in which I lecture about SLA theory and research (telling immensely entertaining stories, of course!); I have asked students to read increasingly fat books about SLA theory and research; and I have led critical small group discussions of these ideas in my classes. I have been teaching my students to talk about SLA, and then telling them, ‘So now you can talk about SLA, go apply what you've learned in your teaching!’ That's a huge disconnect! I was not scaffolding my students in actually doing SLA, learning to analyze learner language for themselves, for their own purposes. I have come to believe that my introductory SLA course was out of balance, with too much ‘Lecture’, and not enough ‘Lab’. It was way too top-down, way too much about me and what I thought, and needed more hands-on activities to engage teachers and beginning researchers in actually doing learner language analysis for themselves, reporting what they saw, and reflecting together on the implications of their own analysis for their own practice.

So what I have begun to believe is that what I do in SLA could use more border-crossing into new ways of teaching and into non-traditional language learning communities. I suspect it is not just me. I suspect that as a field, we value and reward our campus-bound SLA theoreticians and researchers more than SLA practitioners who cross borders to gather data, to apply theories, to learn, to engage in reflective practice, and to provide the kind of two-way exchange of information the field needs between theoreticians/researchers and excellent local reflective practice.

Around the year 2000, because colleagues challenged me, I decided to try to cross a couple of borders in my own research and my own SLA teaching. The record and outcome of these efforts in both SLA research and SLA teaching can be found in two books that kindred spirits at Oxford University Press were kind enough to publish in 2009.

The books represent two efforts at border-crossing. In the first (Tarone & Swierzbin Reference Tarone, Bigelow and Hansen2009), my colleague Bonnie Swierzbin and I tried to change the way we were teaching an introductory SLA course to classes of both beginning researchers and L2 teachers. I have already described the one-way transmission model I had used over two or more decades in my class. But Freeman & Johnson (Reference Freeman and Johnson1998) made a strong case that such courses in SLA were simply not a useful part of the knowledge base needed by L2 teachers and so should no longer be required of them. “What?”, I reacted, “How could that be?” But then I took a second look. Bonnie and I decided to work together to write a curriculum to help us move away from the one-way transmission model – a lecture course – as a way to deliver facts about SLA theories and theoreticians to teachers and beginning researchers. Instead, we would design a course, and eventually a book (Tarone & Swierzbin Reference Tarone, Bigelow and Hansen2009), that would use a bottom-up, inductive approach where our students would ‘Learn By Doing’. We would introduce our SLA class to learner language directly, showing them videos of different learners all doing the same communication tasks, and then engaging them in a process of exploration and discovery about the nature of the learner language they saw. We would ask them, as they did so, to reflect, to think about what patterns they saw in learner language and why they thought the patterns were occurring. We would relate these thoughts to various theories only when they were relevant. I remember telling Roy Lyster we were planning to write a radically different, more hands-on book on SLA for teachers; “That'll be really difficult!”, he said. He was prescient. Bonnie and I spent over a year video-recording six learners of ESL with different native languages, at different proficiency levels, all doing the same unrehearsed communication tasks; we transcribed their speech and analyzed it, and then created activities for teachers and beginning researchers to use that would lead them through a process of exploration and critical thinking about the nature of learner language and of SLA itself. We explicitly adopted the framework of Dick Allwright's Exploratory Practice (Allwright Reference Allwright2003; Allwright & Hanks Reference Allwright and Hanks2009), using videos and questions prompting our students to think about what was happening in the learner language they saw. Because we posed questions to teachers in our book, the publisher insisted we put ‘answers’ in the back; but because the practicing teachers in our classes had used their language classroom experience to come up with possible interpretations of the language samples that we never would have imagined, we often ended up with several possible answers for each question.

Our book was published in 2009. I knew we had crossed some kind of Rubicon when one of the sales representatives staffing the OUP booth the year the book came out greeted me with the question: ‘What the hell is this?’ It certainly didn't look like any of the other SLA books. But it has worked well for us in our classes and workshops. As we used these learner language materials with teachers, we learned that teachers were looking for things and seeing things in learner language that we, and other SLA researchers, were not. For example, teachers in our workshops seemed to be far more interested than SLA researchers in the syntactic complexity that appeared in learner language when it was used for critical thinking in academic work. Though our book has yet to be reviewed by a major journal, crossing this conceptual border has given us lots of local discoveries and led us into new areas of research we would never have ventured into. A grant from the U.S. Department of Education Title VI International Research and Studies Program enabled us to use the same approach at CARLA to develop online video materials focused on Chinese, Japanese, Korean, and Persian learner language. Using exactly the same methodology, CARLA's research team video-recorded learners of these Asian languages carrying out the same unrehearsed communication tasks. In using these materials in teacher workshops, we learned that simply showcasing learners’ unrehearsed communication in these languages was a border-crossing event for teachers who had previously never even considered it possible for third-semester learners of their languages to do that. And so we all continue to learn together.

The second book records our efforts to cross borders in carrying out SLA research with nontraditional populations. In about 2000, Martha Bigelow, then a new assistant professor in another college, crossed a disciplinary border in just coming over to our college to propose collaborative research. We discussed the growing numbers of Somali immigrants, many of them with little literacy in their native language, who were moving into the Cedar Riverside housing area just a couple of blocks from the edge of our campus. Some of our graduate students were teaching them in Adult Basic Education programs. Martha and I decided to try to replicate a mainstream SLA study, not with the usual participants drawn from the university community, but with some members of the Somali immigrant community. Martha began building relationships at Cedar Riverside, getting acquainted, supporting a soccer team, and so on. We applied for – but didn't get – a National Science Foundation grant, but a smaller one enabled us to hire Kit Hansen as a research assistant. As a team, we designed a replication study on recasts, and Kit and Martha crossed multiple borders to gather the data: geographical borders were easier; harder ones were cultural, religious, and role-related (we had to negotiate all kinds of cultural assumptions in ourselves and others, not the least of which involved issues of trust across communities, including our own). Martha published a monograph (Bigelow Reference Bigelow2010) to record her learning and advocacy. We wrote and published papers that explicitly laid out ways the results of the recast study challenged current SLA theory; Bigelow, delMas, Hansen & Tarone (2006) even won the 2007 TESOL/Heinle Distinguished Research Award. The back cover of our book quotes George Yule, who states our claim clearly: ‘the points made have potentially theory-shattering consequences for some of the most fashionable notions current in SLA’. Six years later, those consequences remain only ‘potential’; no SLA theorist has yet responded in any substantive way to the findings and claims in this volume. Our simple act of crossing borders by carrying out an SLA study with a decidedly nontraditional population, even when it produced results that we think are central to the SLA research enterprise, and very important for educating those same populations, turned out to be surprisingly easy for theoreticians in our field to ignore. We have a way to go to achieve a good two-way flow of communication from the field into theory and back again.

Although the rewards of border crossing are clearly of different types, I want to conclude from our experience that more applied linguists should consider border-crossings of their choice. Borders can have their functions and I do not have the energy to try to cross all of them at once. I think we need to choose our exploratory adventures strategically. But I also think that as SLA researchers, and indeed applied linguists in general, we will all benefit from a little more border crossing. For myself, I want to continue to move away from traditional ‘Lecture’ styles of teaching SLA, and I want to encourage my students and myself to gather data from at least some language learners who are not highly educated or literate. When I consider theories of SLA, I find myself favoring theories that might help me gain insight into language puzzles and problems in the lives of all kinds of people, from all kinds of communities, from all parts of the world, whether outside or inside of educational institutions. That is not to say I am not also an applied linguist who works inside the academy; we need good theories and good research. But I do think we need more border-crossing – both in the way we teach others to know and do SLA research, and in the way we conduct SLA research.

To sum up, I would like to propose five principles to achieve a better balance in what we do in the field of SLA:

1. 1. Do not shore up SLA theory at the expense of good learner language data. Build a system whereby learner language data and questions from L2 classrooms from much more diverse international contexts can be published just as medical case studies are, and make sure those data and questions are seriously considered by SLA theorists, and have a serious impact upon their theories. It should not be possible for theorists to just ignore data that is inconvenient to their theory, or to dismiss it because it is ‘just one case’ or because it was produced off campus.

2. 2. Seek feedback on the assumptions of SLA researchers and theorists from those who study multilingual learners in many other parts of world; make sure that feedback includes good data on learner language obtained from every corner of human society. In particular, include learner language data obtained from non-literate and low-literate multilingual adults, and from multilingual language learners in developing world contexts. Do not just ignore what is published about the Vaupes tribe in Brazil; SLA researchers should go there to gather good data on their acquisition processes, to determine whether and to what extent these individuals acquire second languages fits or contradicts current theory.

3. 3. Supplement the predominantly ‘Lecture’ focus of introductory SLA classes for L2 teachers and for beginning graduate student researchers; move away from a one-way transmission model focused on discussing theories. Add more ‘Lab’ activities to help SLA classes learn by doing; show them samples of learner language paired with hands-on exploratory activities and questions that can teach them how to look at learner language for themselves (you can find some in Tarone & Swierzbin Reference Tarone and Swierzbin2009, and at the free CARLA Learner Language website). Teach teachers how to do the kind of critical analysis of learner language and learning processes that can confront the SLA researcher and theorist's core assumptions, and test the theory in the crucible of real-world, local learning contexts.

4. 4. Support hands-on, inquiry-based in-service teacher development such as Exploratory Practice (Allwright Reference Allwright2003; Allwright & Hanks Reference Allwright and Hanks2009), whereby L2 teachers and their students learn to use the tools of action research to improve life in their classrooms by asking and answering their own questions. In Exploratory Practice, teachers and students learn to identify and explore their own puzzles as they crop up in their language learning and teaching classrooms, and do so in a way that makes everyone's life better. Their findings can and should be relayed to SLA researchers and theoreticians in academia. Tell publishers you want more inquiry-based, hands-on books on action research and Exploratory Practice to counterbalance books disseminating theory and published research findings.

5. 5. Value and model advocacy, on behalf both of multilingual learners and of the highest levels possible of language teacher education. I have learned from my role models and from my own students how important it is, not just to build such programs and teach in them, but also to advocate for their very existence and to try to teach administrators and colleagues why these programs are so very important in building a multilingual, multicultural university, school system, city, and society.

I'll conclude with a short story about one of the many graduates of the University of Minnesota's M.A. ESL program who are model border-crossers. Randall Gwin earned his M.A. ESL in 1998 and then went with his young family to live in Kyrgyzstan, where he still lives and teaches English and English teachers. He's now written an EFL methods book (Gwin Reference Gwin2009) and works with a group, ‘A Way To Connect’, that includes local experts as well as other alums of our M.A. ESL program (http://awaytoc.com/awaytoc-directors-and-advisors.html). The group has ties to work in Russia, Uzbekistan, Costa Rica, Vietnam, and other sites. They post the following principles on their website: http://awaytoc.com/about-us/:

1. a. there are no unimportant individuals, all are worthy of respect;

2. b. learning is life-long and we are all learners together, learning by doing;

3. c. there are no unimportant groups, and all deserve respect;

4. d. we should have high standards and keep our word;

5. e. we strive to create resources not deplete them, and we succeed when our students surpass us.

To the extent that Randall Gwin and his border-crossing group have learned from me, and to the extent that I now try to emulate them, I believe I have succeeded, because my students surpass me.